Genomic integrity depends critically on the fidelity and efficiency of DNA replication. Processive polymerases can stall at DNA damage sites and translesion synthesis is then dominated by bypass polymerases, involving error-free (mutation-avoiding) or error-prone (mutation-generating) pathways. Formation of oxidative damage sites are accelarated under conditions of smoking-induced oxidative stress, and markedly increased levels of oxidative damage adducts have been detected in lung tissue that has been chronically exposed to cigarette smoke. Our goal is to understand the molecular interactions that define the mutagenic spectrum associated with replication of oxidative damage sites by bypass polymerases. Our proposed crystallographic studies will be undertaken on the most prevalent oxidative damage lesions, namely, 8-oxoguanine, the stable ring-opened 5-guanidino-4-nitroimidazole adduct and the fused bicyclic spiroiminodihydantoin adduct, positioned at template-primer junctions, as part of binary and ternary (with incoming nucleoside triphosphates) complexes, with the thermophilic Dpo4 bypass polymerase. We have already solved crystal structures of binary and ternary complexes of oxoG- containing template-primer junction, Dpo4 and incoming dCTP that have provided detailed insights into dCTP-binding and dCTP-incorporation steps, and following additional planned experiments, elongation steps. Our proposed structural studies of Dpo4 ternary complexes of oxidative guanine adducts in the C- *G-C sequence context will be extended to the C-*G-T sequence context, which constitute a sub-set of mutational 'hot spots' in the p53 tumor suppressor gene. Our efforts should elucidate the geometric fit, alignment and register for individual oxidative damage lesions of varying size and shape positioned in the active site of Dpo4, should determine the specific interactions and pairings of the lesion site with complementary and non-cognate incoming nucleoside triphosphates, and should identify key residues and alignments for facilitating the divalent cation-mediated nucleotidyl transfer reaction. The proposed studies should provide structural insights into how bypass is modulated by lesion architecture and base sequence context, and provide explanations for the distribution of point mutations relative to frameshift deletions.